The etiology of endometrial cancer is relatively well understood. Estrogen stimulation of the endometrium without the modulatory effects of progestins is the major cause. Estrogen replacement therapy (ERT) in menopause and obesity are the principal risk factors. The effect of the latter is probably due to the association between postmenopausal obesity and circulating bioavailable estrogen levels. Oral contraceptives and pregnancy, both of which deliver estrogen stimulation to the endometrium but with the continuous modulatory influence of progestins, are associated with reduction in risk. Combination hormone replacement therapy in which a progestin is added to estrogen for all or part of the monthly cycle results in no increase in endometrial cancer risk over that of a non-user of hormone replacement. Despite the fact that ERT and obesity are the major risk factors, only a small proportion of women using ERT or even with extreme obesity will develop endometrial cancer. It would be important from a public health as well as from a mechanistic view to be able to predict which women those will be. We propose to evaluate a series of eight candidate genes (CYP17, CYP19, HSD17B1, ER, CYP1A1, CYP1B1, COMT andPR) in the estrogen biosynthesis, transactivation and metabolism pathways to determine if the effects of these risk factors might be mediated or modified by genetic variability. We will evaluate this question in the context of a prospective epidemiologic study of 133,000 female California teachers (the California Teachers Study) using a nested case-control design. We will also examine in detail the possible impact of phytoestrogens on endometrial cancer risk reduction in conjunction with HRT, obesity and the eight candidate genes under evaluation.